GeForce GTX 1070 and Radeon HD 5970 - video card comparison
Passmark - benchmark score GeForce GTX 1070 vs Radeon HD 5970
GeForce GTX 1070
13916
Radeon HD 5970
2350
GeForce GTX 1070
Radeon HD 5970
Architecture
Pascal (2016−2021)
TeraScale 2 (2009−2015)
GPU code name
Pascal GP104
Hemlock
Market segment
Desktop
Desktop
Release date
May 2016
November 2009
Bus support
PCIe 3.0
-
Interface
PCIe 3.0 x16
PCIe 2.0 x16
Recommended system power (PSU)
500 Watt
-
Supplementary power connectors
8-pin
1x 6-pin + 1x 8-pin
SLI options
+
-
Boost clock speed
1683 MHz
-
Core clock speed
1506 MHz
725 MHz
CUDA cores
1920
-
Floating-point performance
6,463 gflops
2x 2,320.0 gflops
Manufacturing process technology
16 nm
40 nm
Number of transistors
7,200 million
2,154 million
Pipelines / CUDA cores
1920
1600
Power consumption (TDP)
150 Watt
294 Watt
Maximum GPU temperature
94 °C
-
Texture fill rate
202.0
58.00
Memory type
GDDR5
GDDR5
Maximum RAM amount
8 GB
1 GB
Memory bus width
256 Bit
256 Bit
Memory clock speed
8 GB/s
4000 MHz
Memory bandwidth
256 GB/s
128.0 GB/s
Shared memory
-
-
Display Connectors
DP 1.42, HDMI 2.0b, Dual Link-DVI
2x DVI, 1x mini-DisplayPort
Multi monitor support
+
-
HDMI
+
-
G-SYNC support
+
-
DirectX
12 (12_1)
11.2 (11_0)
CUDA
+
-
OpenCL
1.2
1.2
OpenGL
4.5
4.4
Shader Model
6.4
5.0
Vulkan
+
N/A
Ansel
+
-
GPU Boost
3.0
-
VR Ready
+
-